United States: Complications Associated with Regional Anesthesia (An American Society of Anesthesiologists’ Closed Claims Analysis)


Neuraxial n = 317

PNB n = 128

p value

Total n = 445

Sex of patients


236 (74 %)

61 (48 %)

p < 0.001

297 (67 %)

Age in years (n = 435)

 Mean (SD)

46 (19)

52 (15)

p = 0.001

48 (18)

 Pediatric (16 and under)

2 (1 %)

2 (2 %)

p = 0.326

4 (1 %)

ASA physical status (n = 432)

 ASA 1–2

181 (58 %)

92 (76 %)

p = 0.001

273 (63 %)

Emergency (n = 441)


58 (18 %)

6 (5 %)

p < 0.001

64 (15 %)

Procedure scheduled as inpatient/outpatient (n = 435)


287 (92 %)

30 (24 %)

p < 0.001

317 (73 %)


25 (8 %)

93 (76 %)

118 (27 %)

Category of anesthetic care


112 (35 %)

73 (57 %)

p < 0.001

185 (42 %)


142 (45 %)

0 (0 %)

142 (32 %)

 Acute pain

63 (20 %)

55 (43 %)

118 (27 %)

PNB peripheral nerve blocks, ASA American Society of Anesthesiologists

N = 445 unless otherwise specified. Claims with missing data excluded

p-values for differences between neuraxial and peripheral nerve block groups by Fisher’s exact test for proportions and t-test for age

The specific distribution of the blocks associated with claims is summarized in Table 32.2. The most common neuraxial blocks in claims were lumbar epidural (44 % of neuraxial block claims) and spinal (36 %). Among peripheral nerve blocks, the most common were interscalene (45 % of peripheral block claims), followed by eye blocks (14 %), axillary blocks (12 %), and femoral blocks (12 %). It is difficult to know whether the frequency of the appearance of the individual block types mirrors the frequency with which these blocks were performed in clinical practice, but there are some indications that it does not. A national survey of practice patterns for peripheral nerve blocks published in 1998 indicated that anesthesiologists reported performing the following upper extremity blocks in decreasing order of frequency: axillary, interscalene, wrist, supraclavicular, elbow, infraclavicular while for the lower extremity the block order in descending frequency was ankle, femoral sciatic, popliteal [8]. If this study accurately reflects national practice and the pattern has not changed greatly over the ensuing years, it would suggest that interscalene block is a potentially higher risk block for injuries with malpractice claims, while ankle blocks appear to be lower risk relative to the frequency with which they are performed. The same consideration of malpractice risk relative to frequency of clinical volume may apply to eye blocks but it is even more difficult to capture data on the number of these blocks performed by anesthesiologists.

Table 32.2
Types of regional blocks

Neuraxial blocks (n = 317)

n (% of 317)

 Lumbar epidural

140 (44 %)


113 (36 %)

 Spinal-lumbar epidural combination

31 (10 %)

 Thoracic epidural

26 (8 %)

 Caudal epidural

1 (0 %)

 Cervical epidural

1 (0 %)

 Unspecified epidural

5 (2 %)

Peripheral nerve blocks (n = 128)

n (% of 128)


58 (45 %)

 Eye blocks

18 (14 %)


9 (7 %)


7 (5 %)

  Eye block, unspecified

2 (2 %)


15 (12 %)


15 (12 %)


7 (5 %)


5 (4 %)


3 (2 %)


2 (2 %)

 Bier block

2 (2 %)


1 (1 %)

 Multiple techniques

2 (2 %)

A comparison (Fig. 32.1) of the damaging events between peripheral nerve blocks and neuraxial blocks indicates that, while regional anesthesia-related events were most common in both groups, there were proportionately more cardiovascular and respiratory damaging events in the neuraxial group than the peripheral nerve block claims. This finding may be related to the higher proportion of ASA 3–5 patients in the neuraxial group or due to the cardiovascular effects of a neuraxial block. Most (61 %) peripheral nerve block claims involved temporary or nondisabling injuries, while neuraxial claims had proportionately more permanent disabling injuries (36 % of neuraxial claims vs. 27 % peripheral nerve blocks) and death (28 % neuraxial vs. 12 % peripheral nerve blocks, Fig. 32.2). Only one-third (36 %) of neuraxial block claims had temporary injuries, which were predominately associated with obstetric claims (69 %).


Fig. 32.1
Primary damaging events in neuraxial vs. peripheral nerve blocks. PNB peripheral nerve blocks, CV cardiovascular events. Events differed between groups (p = 0.001 by Fisher’s exact test ). Neuraxial blocks were associated with proportionately more cardiovascular and respiratory events


Fig. 32.2
Severity of injury in neuraxial vs. peripheral nerve blocks. PNB peripheral nerve blocks. Death was more commonly associated with neuraxial blocks, while temporary injuries were more commonly associated with PNB. Severity differed between groups (p < 0.001 by Fisher’s exact test)

Claims reviewers assessed the anesthesia care as appropriate in about two-thirds of both neuraxial (62 %) and peripheral nerve block claims (65 %, Table 32.3). Approximately half of all regional anesthesia claims resulted in payment. Payments for neuraxial blocks were significantly higher than for peripheral nerve blocks (median payment $503,000 vs. $248,000, respectively, p = 0.001, Table 32.3), which is likely related to the higher severity of injury associated with neuraxial blocks.

Table 32.3



p value


Appropriate anesthesia care (n = 405)

177 (62 %)

77 (65 %)

p = 0.652

254 (63 %)

Claim paid (n = 442)

172 (55 %)

57 (45 %)

p = 0.074

229 (52 %)

Payment amount (n = 229)

 Median (p < 0.01)



p < 0.001


 Interquartile range



PNB Peripheral nerve block

N = 445 unless otherwise specified. Claims with missing data excluded

Payment amounts adjusted to 2014 dollars by CPI and rounded to nearest $1000.00

Payment median p-value by Mann–Whitney U Test between neuraxial and PNB groups

Appropriateness and claim paid p-values for differences between neuraxial and PNB groups by Fisher’s exact test for proportions

Neuraxial Cardiac Arrest

The problem of neuraxial cardiac arrest occupies a central place in the history of the Closed Claims Project as it was the subject of the first article and review of the database published in 1988 [2]. In another review of closed claims of regional anesthesia complications encompassing the 1980s and 1990s, there were 73 neuraxial cardiac arrest claims making it at that time the largest category of block-related regional anesthesia claims with death or brain damage [9]. In contrast, the review for this chapter covering events that occurred in the year 2000 or later contained 12 claims for neuraxial cardiac arrest in 13 years with two of those claims being severe bradycardic episodes that responded to pharmacologic treatment without progression to full arrest or requirement for chest compressions. Only one of the 12 claims in the recent review fit the most common pattern described in 1988 by Caplan et al. [2], that is, of an arrest occurring approximately 30 min after initiation of the anesthetic in a patient receiving moderate to heavy sedation. Three of the more recent claims were for arrests occurring during epidural anesthesia, while nine claims were associated with spinal anesthesia, one of which was due to the spinal component of a combined spinal epidural. In three claims, the arrest occurred within the first 5 min after the spinal was placed and in all of these cases the resuscitation was unsuccessful with death as the final outcome. The timing of the arrest relative to the initiation of the anesthetic in the remaining claims was quite variable with one occurring over 1 h into the procedure and another coinciding with placement of the dressing at the end of surgery. There were four deaths in the 12 claims and five patients had hypoxic–ischemic brain injury ranging from mild residual cognitive deficits to severe, disabling encephalopathy. It is likely that most episodes of bradycardia occurring during neuraxial anesthesia respond to single doses of ephedrine or anticholinergic medications, but in these claims there was a pattern of failure to respond to multiple repeated doses of these medications, raising the possibility that clinicians may need to consider an earlier transition to epinephrine when faced with bradycardia resistant to standard treatments. Six of the claims occurred in the setting of obstetric anesthesia, two of these arrests resulted in the death of the patients, two were associated with hypoxic –ischemic brain injury, and the other two had severe bradycardia without cardiac arrest and no long-term sequelae. The role of establishing and maintaining good rapport with patients is highlighted by these two claims as the records suggested that preexisting poor rapport was possibly a more significant trigger for the claim than the anxiety and discomfort that accompanied the brief, successfully treated arrhythmia. Payments for damages were made in six claims with a median payment of $383,250 (range $187,500–$2,211,000).

Two frequently suggested mechanisms to explain neuraxial cardiac arrest are as follows: (1) left ventricle hypovolemia causing a paradoxical bradycardic response via stretch/mechanoreceptors (the Bezold–Jarisch reflex ); and (2) blockade of the cardiac accelerator fibers with high sympathetic blockade >T4 [10]. Baseline bradycardia and male gender have been identified as risk factors for severe bradycardia (<40 beats per minute) under neuraxial blockade, but these were not present in the claims in this review. Previous studies are consistent with the findings of this review in that bradycardia which may be a premonitory sign of impending arrest can be widely distributed throughout the time course of neuraxial anesthetics [11].

Although timely initiation of appropriate treatment for neuraxial cardiac arrest has been associated with full recovery in several case studies, many patients are refractory to rescue [12]. This refractoriness to treatment was explored in studies in dogs which demonstrated that the intense sympathetic blockade during spinal anesthesia decreases circulating blood volume and reduces coronary perfusion pressure, thereby rendering cardiopulmonary resuscitation (CPR) ineffective [13]. Moreover, other studies in dogs have shown that neuraxial anesthetic blockade prevents an increase in epinephrine and norepinephrine catecholamine levels during cardiac arrest compared with controls without neuraxial blockade [14]. Consequently, both severe vasodilatation and lack of an appropriate catecholamine response to stress make resuscitation during neuraxial cardiac arrest more difficult.

The apparent dramatic reduction in the number of neuraxial cardiac arrest claims in the database over the past three decades could have a number of explanations, however, given the difficulty of assessing trends in the undefined denominator population, any explanations of this phenomenon should be considered speculative. Several factors could have had an impact on the claims. There may be a reduction in the absolute number of neuraxial anesthetics being performed, particularly for nonobstetric anesthetics, where there has been an increase in peripheral nerve blocks and catheters over the last decade. Changes in anesthetic practices may have reduced the number of neuraxial cardiac arrests relative to the number of neuraxial anesthetics or improved the prevention, recognition, and effectiveness of the rescue therapy provided in neuraxial cardiac arrest. The report by Caplan et al. [2] advocated closer respiratory monitoring when sedation was used with neuraxial anesthesia, the earlier use of epinephrine in the treatment of sudden severe bradycardia with neuraxial anesthesia, and immediate treatment of cardiac arrest with a full resuscitation dose of epinephrine . Whether the dissemination of these ideas or whether other aspects of anesthetic practice contributing to improved patient safety reduced the frequency of malpractice claims for neuraxial cardiac arrest is unclear . Because patients who recover completely from neuraxial cardiac arrests may be less likely to file malpractice claims, the success and failure rates of prompt resuscitation and monitoring cannot be determined from the apparent decrease in the number of claims in the database.

High Block/Total Spinal Block

There are instances where it can be difficult to clearly distinguish, both in the moment or retrospectively, between claims related to neuraxial cardiac arrest and those where cardiac arrest occurred in the setting of unusually high neuraxial block or total spinal anesthesia. As much as possible these claims should be differentiated and reviewed separately as the factors associated with these two complications appear to differ in significant ways. In the most recent review, there were 24 claims involving a high neuraxial block, evenly divided between those arising in obstetric and nonobstetric anesthesia practice.

Of the 12 claims for high block in obstetric practice, four occurred during the provision of labor analgesia while the other eight involved anesthesia for cesarean delivery. All of the claims associated with labor analgesia appeared to have involved unrecognized intrathecal injection. Two of the four labor analgesia obstetric claims involved the questionable practice of injecting local anesthetic through the epidural needle prior to placement of the catheter. The remaining two high blocks associated with labor analgesia cases resulted from injections through the epidural catheter, one with a lidocaine test dose and the other without use of a formal test dose. Among the high blocks associated with anesthesia for cesarean delivery, five occurred with dosing of an epidural and three in the setting of spinal anesthesia. Notable among these cases was an incident where an aspiration test that was negative for cerebrospinal fluid prior to a 10 mL injection of lidocaine was noted to be undeniably positive after it resulted in a total spinal. There is controversy, particularly in obstetric anesthetic practice, regarding what doses of local anesthetic represent appropriate test doses for the identification of an intrathecal catheter when an aspiration test is negative [15, 16]. Data from the limited number of pertinent cases in the database cannot add much clarity to this controversy. High neuraxial blocks occurred both in the setting of dose amounts consistent with historically accepted test doses and in instances where a negative aspiration test apparently gave a false sense of security before an unintended intrathecal injection of a large dose of local anesthetic.

One of the more vexing challenges in obstetric anesthesia arises when the existing labor epidural provides insufficient anesthesia for a cesarean section after injection of what would usually be an adequate dose of local anesthetic. If time permits, one of the options available to the anesthesiologist in this setting is the removal of the epidural and placement of a spinal anesthetic. However, the appropriate spinal local anesthetic dose in this context remains uncertain. In one of the claims in this review, a standard dose was used in a spinal after a failed epidural local anesthetic load, resulting in a high block leading to respiratory and cardiac arrest. In a major review of obstetrical complications, this practice was identified by D’Angelo et al. as having significant potential for dangerously high neuraxial blocks, as 20 % of the high blocks in their review occurred with spinals placed after failed epidurals [17].

In the nonobstetric setting there were six high blocks resulting from the placement of thoracic epidurals for analgesia, three claims in the setting of spinal anesthetics, three from lumbar epidurals, and two from peripheral nerve blocks (an interscalene and a paravertebral). The two high thoracic epidural blocks that were due to unrecognized intrathecal placement left the patients with residual neurologic injury to the spinal cord, a syrinx in one claim and cauda equina syndrome in the other. The importance of recommendations to avoid placing neuraxial blocks in adults who are unresponsive or poorly responsive due to sedation or general anesthesia is raised in this group of claims, as the syrinx of the spinal cord occurred in a patient whose epidural was placed under general anesthesia. The remaining high thoracic epidural blocks occurred largely in the setting of aggressive use of lidocaine boluses to rapidly achieve analgesia in postoperative and posttrauma patients who were vulnerable to the effects of a high thoracic sympathectomy. From the lack of intensity in the monitoring during the immediate postblock period, it appeared that the providers were unaware of and unprepared for the effect that a large dose of lidocaine could have when delivered as a bolus to the thoracic epidural space. In many of the high block claims, the combination of delayed recognition of a high block and the inability to simultaneously address both the apnea and hypotension accompanying this complication caused severe patient injuries. In some instances, delayed recognition was due to the fact that the anesthesiologist administered a dose of local anesthetic and then very shortly after that left the patient to attend to other tasks. The anesthesiologist was then belatedly aware of the problem only upon return from the distracting task or when alerted by nursing staff.

Only gold members can continue reading. Log In or Register to continue

Oct 25, 2017 | Posted by in Uncategorized | Comments Off on United States: Complications Associated with Regional Anesthesia (An American Society of Anesthesiologists’ Closed Claims Analysis)

Full access? Get Clinical Tree

Get Clinical Tree app for offline access